Comparative study of the synthesis of layered transition metal molybdates

Article abstract of DOI:10.1016/j.jssc.2009.10.011

Mixed metal oxides (MMOs) prepared by the mild thermal decomposition of layered double hydroxides (LDHs) differ in their reactivity on exposure to aqueous molybdate containing solutions. In this study, we investigate the reactivity of some T-Al containing MMOs (T=Co, Ni, Cu or Zn) towards the formation of layered transition metal molybdates (LTMs) possessing the general formula AT₂(OH)(MoO₄)₂·H₂O, where A=NH₄⁺, Na⁺ or K⁺. The phase selectivity of the reaction was studied with respect to the source of molybdate, the ratio of T to Mo and the reaction pH. LTMs were obtained on reaction of Cu-Al and Zn-Al containing MMOs with aqueous solutions of ammonium heptamolybdate. Rehydration of these oxides in the presence of sodium or potassium molybdate yielded a rehydrated LDH phase as the only crystalline product. The LTM products obtained by the rehydration of MMO precursors were compared with LTMs prepared by direct precipitation from the metal salts in order to study the influence of preparative route on their chemical and physical properties. Differences were noted in the composition, morphology and thermal properties of the resulting products. Crown Copyright

Full text of DOI:10.1016/j.jssc.2009.10.011

ARTICLE IN PRESS
Journal of Solid State Chemistry 183 (2010) 198–207
Contents lists available at ScienceDirect
Journal of Solid State Chemistry
journal homepage: www.elsevier.com/locate/jssc
Comparative study of the synthesis of layered transition metal molybdates
1
Ã
S. Mitchell, A. G o´ mez-Avil e´ s , C. Gardner, W. Jones
Department of Chemistry, The University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Mixed metal oxides (MMOs) prepared by the mild thermal decomposition of layered double hydroxides
(LDHs) differ in their reactivity on exposure to aqueous molybdate containing solutions. In this study,
we investigate the reactivity of some T–Al containing MMOs (T=Co, Ni, Cu or Zn) towards the formation
Received 11 August 2009
Received in revised form
1
4 October 2009
of layered transition metal molybdates (LTMs) possessing the general formula AT
2
(OH)(MoO
4
)
2
ꢀ H O,
2
Accepted 15 October 2009
Available online 28 October 2009
+
4
+
+
where A=NH
, Na or K . The phase selectivity of the reaction was studied with respect to the source of
molybdate, the ratio of T to Mo and the reaction pH. LTMs were obtained on reaction of Cu–Al and Zn–Al
containing MMOs with aqueous solutions of ammonium heptamolybdate. Rehydration of these oxides
in the presence of sodium or potassium molybdate yielded a rehydrated LDH phase as the only
crystalline product. The LTM products obtained by the rehydration of MMO precursors were compared
with LTMs prepared by direct precipitation from the metal salts in order to study the influence of
preparative route on their chemical and physical properties. Differences were noted in the composition,
morphology and thermal properties of the resulting products.
Keywords:
Layered double hydroxides
Mixed metal oxides
Transition metal molybdates
Calcination and rehydration
Crown Copyright & 2009 Published by Elsevier Inc. All rights reserved.
1
. Introduction
Molybdenum containing compounds are important in the field
[11]. Table 1 reports the LTMs of this series which, to our
knowledge, are known to exist. Two structural polytypes have
been identified which were classified by Pezerat as the
phases, respectively [25]. Simulated structures and PXRD patterns
of these phases are shown in Fig. 1. In the structure the layers
F
x
and
F
y
of catalysis as they are known to be catalytically active for several
processes including oxidative dehydrogenation (e.g. of simple
alkanes [1,2]), hydrodesulfurization (e.g. of petroleum [3]) and
selective oxidation (e.g. the epoxidation of alkenes [4] or alkyl
olefins [5] or for the selective oxidation of alcohols [6]). The
catalytic properties of transition metal molybdates (TMMs)
containing first row transition metals have been studied where
the transition metal(s) may also play a catalytic role. As transition
metals frequently exist in more than one stable oxidation state,
transition metal containing compounds often exhibit porosity due
to the high structural flexibility of their constituents [7,8]. In
addition to their catalytic properties, TMMs also attract attention
due to the diverse and potentially tuneable, optical, electronic and
magnetic properties which they may possess [9,10].
F
x
are composed of chains of edge sharing transition metal octahedra
which are connected by tetrahedral molybdate groups [24]. In the
y
F structure transition metal octahedra edge share to form sheets.
In this case, the molybdate tetrahedra cap an ordered cation
vacancy [19]. The layers stack in the c direction and are held
together by electrostatic and hydrogen bonding interactions. The
layers are negatively charged, electroneutrality being maintained
+
by the incorporation of charge balancing cations (A ) within the
interlayer space. Whereas the
F
y
x
structure has been obtained
polytype has to date only been
+
+
+
+
with A =Na , K or NH
4
, the
F
+
+
4
reported for A =NH .
The synthesis of LTMs has not been widely studied. Many of
the LTMs reported to date were isolated by authors investigating
the reaction between different transition metal salts with alkali
molybdate species. In general their synthesis was not specifically
targeted but rather the intention was to prepare novel mixed
metal materials with potentially interesting (catalytically active,
porous, etc.) properties [15]. An alternative synthetic route for the
It is possible to synthesise layered TMMs (LTMs) of general
2
+
2+
2+
2+
2+
formula AT
2
2+
OH(MoO
4
)2(H
2
O) where T =Mn , Fe , Co , Ni
,
,
2
+
+
+
4
Cu or Zn (or a combination of these metals) and A =NH
+
+
Na or K . Layered materials are of particular interest for their
intercalation chemistry, large potentially accessible internal sur-
face area and as precursors for the generation of 2D nano-sheets
preparation of LTMs of the
y
F structure which involved the
rehydration of a mixed metal oxide (MMO), derived from a T–Al
layered double hydroxide (LDH) [26,27] precursor, in an aqueous
ammonium heptamolybdate solution was introduced by Levin
Ã
Corresponding author. Fax: +44 1223 336017.
E-mail address: wj10@cam.ac.uk (W. Jones).
2
+
2+
1
et al. [23]. LTMs were successfully obtained with T =Zn or a
Visiting from: Instituto de Ciencia de Materiales, CSIC, Cantoblanco 28049,
2
+
2+
2+
2+
2+
Madrid, Spain.
mixture of Zn and Cu or with Co or Ni but not with Mg
0022-4596/$ - see front matter Crown Copyright & 2009 Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2009.10.011

ARTICLE IN PRESS
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
199
2
+
or Ni
as the only divalent metal. This method provides an
terms of their composition, morphology and thermal properties.
As the properties of MMOs derived from layered precursors may
be influenced by the preparative route of the layered precursor
[30], we also study the influence of precursor preparation route
on the LTMs obtained.
alternative, low temperature, environmentally friendly (e.g. due
to the elimination of NO
In this study, we report the reaction of T–Al MMOs where
T=Co, Ni, Cu or Zn with aqueous solutions containing ammonium
heptamolybdate or sodium or potassium molybdate. The extent of
ꢁ
3
waste streams) preparation route.
2
4
ꢁ
6ꢁ
molybdate polymerisation in aqueous solution (MoO , Mo
7
O
,
24
2. Experimental section
etc.) is known to be influenced by both the solution pH and
concentration of molybdate species present [28]. The solution pH
also influences the extent of LDH dissolution [29]. For these
reasons we have investigated the influence of reaction conditions
including the ratio of T to Mo, reaction pH and the LDH precursor
crystallite size on the reactivity of the MMOs. We compare the
LTM products prepared by the rehydration method with analo-
gous LTMs synthesised via the standard route of precipitation in
2.1. Chemicals
All chemicals used were reagent grade, used as supplied by
Sigma–Aldrich.
2.2. MMO method
Precursor LDHs were prepared both by the commonly reported
route of coprecipitation [26,27] and by the urea hydrolysis
method, a method which is known to yield LDH products with a
higher aspect ratio [31]. In the coprecipitation synthesis, a 50 ml
Table 1
2
+
+
Summary of T and A combinations and observed structure of LTMs previously
reported of general formula AT
2
OH(MoO
4
)
2
ꢀ H
2
O.
2
+
+
T
A
mixed metal salt solution containing T(NO
)
ꢀ xH
and Al(NO
0.01 mol) and an aqueous solution containing NaOH (2 M) and
Na CO (2 M) were prepared. These solutions were combined by
O (0.02 mol),
3
2
2
2
+
2+
2+
2+
2+
where
T
=Zn
,
Cu
,
Ni
or Co
,
3
)
3
ꢀ 9H
2
O
+
4
+
+
NH
Na
K
(
2
+
+
Mn
F
–
x
[12]
F
F
F
F
–
x
x
x
x
[13]
F
x
F
x
F
x
F
x
–
[14]
[16]
[17]
[17]
2
3
2
Fe
[15]
dropwise additions to distilled water (50 ml) heated at 55 1C, at
relative rates such that a constant pH of 9.070.1 was maintained.
Following precipitation the solution was then aged for 24 h at
70 1C. The resulting precipitates were collected by filtration,
washed and dried. For the urea hydrolysis method, a 50 ml metal
salts and urea solution was prepared by dissolving
2
+
a
Co
F
F
F
F
x
y
y
y
[17]
[19]
[22]
[17,18]
[20,21]
2
+
Ni
2
+
+
Cu
2
Zn
[21,23]
F
x
[17,24]
F
x
[17]
a
Reported by Pezerat et al. to have the
compound with the structure.
x
F structure. We have synthesised this
F
y
Zn(NO
)
2
ꢀ 6H
O
(0.666 mol) with Al(NO
)
2
ꢀ 9H
2
O (0.333 mol)
3
2
3
Fig. 1. Schematic diagram of (a) the
F
y
and (b) the
F
x
LTM structures of general formula AT
2
OH(MoO
4
)
2
ꢀ H
2
O, projected along the (110) plane. The simulated PXRD
patterns for these structures (c) are also shown.

ARTICLE IN PRESS
2
00
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
2
+
3+
ꢁ3
and urea such that [T ]+[Al ]=1 mol dm
(
and [urea]/
thermal gravimetric (DTG) curves were obtained by heating the
2
+
3+
[T ]+[Al ])=3.3 in distilled water. The solution was then
heated at 110 1C under reflux for 24 h. The resulting precipitate
was collected, washed and dried.
sample (10–20 mg) from 25 to 800 1C in air, at a ramp rate of
ꢁ
1
10 1C min . In cases where evidence of sample oxidation was
observed (i.e. for samples containing T=Cu) the measurements
ꢁ
1
Formation of MMOs was achieved by calcination of the
2
were repeated under a flow of N gas (30 ml min ).
precursor LDHs in
a
carbolite furnace at 500 1C for 3.5 h.
Scanning electron microscopy (SEM) was undertaken using a
JEOL 5800LV instrument. Samples were dispersed from ethanol
onto a freshly cleaved Mica surface. A thin platinum coating was
applied to reduce the effects of sample charging.
Calcination temperatures of 350, 250 and 200 1C were also used
when the MMO obtained at 500 1C was found to be unreactive
during the rehydration step.
A typical procedure to study MMO reactivity involved reacting
the MMO (0.25 g) with a 50 ml solution of ammonium heptamo-
lybdate or sodium or potassium molybdate ([T]/[Mo]) prepared
with distilled water. The mixture was stirred for 4 h and the
resulting product collected by filtration, washed and dried. The
influence of the T/Mo ratio was studied for MMOs which were
found to be reactive towards the formation of LTMs. Reactions
were undertaken with varying concentrations of ammonium
heptamolybdate solutions (0.2, 0.1, 0.05 and 0.025 M). Similarly
the influence of pH was studied by adjusting the pH after addition
of the MMO precursor to the molybdate solution by addition of
Compositional analysis of C, H and N content was carried our
by the Microanalysis Service at the Department of Chemistry,
University of Cambridge. The T/Mo ratio was also analysed by
energy dispersive X-ray diffraction (EDX) using the JEOL 5800 LV
instrument with an accelerating voltage of 30 KeV. The samples
used for these measurements were not coated with Pt. Results
were averaged over a minimum of 5 sites per sample. Analysis of
the accuracy of the EDX was undertaken by calibration with
samples of known composition. The T/Mo ratio obtained for a
4
ZnMoO sample (Sigma–Aldrich), for example, was found to be
1.0/1.1 with a standard deviation of 70.12.
aqueous solutions of NH
4
OH (35%)/NaOH (1 M) and HNO
3
(1 M)
depending on the source of molybdate used.
3. Results and discussion
2.3. Precipitation method
3.1. Nomenclature
2
+
2+
Here, LTMs were isolated on reaction of MMOs (T =Zn
,
2
+
+
+
Samples prepared during this study are described by the
Cu and A =NH
4
), analogous LTMs were prepared by solution
following nomenclature:
precipitation from the metal salts. It should be noted that LTM
phases may be isolated from solution under a wide range of
conditions, the specific conditions indicated were those used to
obtain the LTMs synthesised for comparative purposes in this
work. Aqueous solutions containing the desired transition metal
T_C_M
T is the transition metal studied Zn, Cu, Ni or Co.
C indicates the nature of the compound: LDH, MMO or LTM.
M indicates the synthesis method: P for precipitation, U for
urea hydrolysis and MP or MU for synthesis by reaction of a
MMO derived from a precursor LDH prepared by
coprecipitation (MP) or by urea hydrolysis (MU).
nitrate (0.007 mol) and (NH
prepared using distilled water (50 ml) with a T/Mo ratio of 1. LTM
4
)
6
Mo
7
O
24 ꢀ 4H
2
O (0.001 mol) were
precipitation was induced by increasing the pH to approximately
6
(for T=Cu) or pH 8 (for T=Zn) by the addition of a basic solution
of NH OH (35%). The resulting solutions were then aged at 60 1C
4
for 4 h. After ageing the precipitates were collected by filtration,
washed with distilled water and oven dried at 100 1C.
3.2. Reaction of MMO: reactivity
2.4. Experimental techniques
Calcination of precursor LDHs at 500 1C led to decomposition of
the layered structure and the formation of a disordered MMO phase,
as expected [23]. PXRD patterns of the resultant MMOs show broad,
Powder X-ray diffraction (PXRD) patterns of the resultant
materials were recorded using a Philips PW3710 diffractometer in
reflection geometry using graphite monochromated CuK radia-
a
tion. Data were collected from finely ground samples pressed on a
flat plate, glass sample holder. Phase identification was achieved
by comparison with PXRD patterns simulated, using the X’Pert
Plus software, from single crystal structural information data;
2 4 2 2 x 3 4 2 2 10
NaZn OH(MoO ) (H O) (F structure) no. 305, H NH Zn Mo O
(F
y
structure) no. 24967, as reported in the International Crystal
Structure Database. Variable temperature data were collected
using an Anton Parr TTK450 low temperature attachment. The
ꢁ
1
sample was heated, at a heating rate of 10 1C min , to the desired
temperature and held at that temperature while the data were
collected. Measurements were performed every 25 1C increment
from 25 to 450 1C.
Fourier Transform Infra-Red (FTIR) spectra were recorded on a
ThermoNicolet Nexus Spectrometer using the Smart Golden Gate
Single Reflection stage Attenuated Total Reflection. A background
spectrum was recorded before the sample was placed on the
sample stage, pressure applied using a sapphire anvil, and data
ꢁ
1
acquired. Data were recorded in the range
n
=550–4000 cm
.
Thermal analysis was performed on a Mettler Toledo SDTA 851
Fig. 2. PXRD patterns of the MMO phases obtained on calcination of their
thermal balance. Thermal gravimetric (TG) and differential
respective precursor LDHs.

ARTICLE IN PRESS
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
201
Table 2
Crystalline phases identified in the T_MMO_P phases obtained following calcination of T_LDH_P precursors to 500 1C, and in the products resulting from their reaction with
aqueous ammonium ðNHþ Þ heptamolybdate or sodium (Na ) or potassium (K ) molybdate (T: MoE1, room temperature).
+
+
4
Precursor
Crystalline phases identified
In MMOs
On reaction ðNH4þ
Þ
+
+
On reaction (Na )
On reaction (K )
Co_MMO_P
Ni_MMO_P
Cu_MMO_P
Co
3
O
4
Co
3
O
4
Co
3
O
4
Co
3
O
4
NiO
CuO
NiO
NiO
NiO
ðCO 32 ꢁÞLDH
ðCO3 Þ LDH
F
F
y
LTM
LTM
ðCO 322 ꢁÞ LDH
ꢁ
2ꢁ
Zn_MMO_P
ZnO
y
ðCO3 Þ LDH
Fig. 3. PXRD and FTIR data of the LDH, MMO precursors and the product obtained on reaction with ammonium heptamolybdate and sodium molybdate solutions for
2
+
T=Zn
.
low intensity reflections, in positions corresponding to those
reported for oxides of the transition metal (T) studied (Fig. 2).
Reflections are indexed according to the crystalline phases identified
molybdate was not incorporated significantly in these products
(Zn/Mo=22 or Cu/Mo=18).
3
+
in Table 2. The location of the Al remains the subject of debate.
Following mild thermal treatment the aluminium is not thought to
phase separate significantly. A higher calcination temperature would
be required to detect the formation of a crystalline aluminium
containing phase e.g. a spinel or oxide phase [32].
3
.3. Reaction of MMO: influence of T/Mo ratio
The formation of LTMs of structure on reaction of
F
y
Zn_MMO_P or Cu_MMO_P precursors with ammonium heptamo-
lybdate was found to be highly dependent on the T/Mo ratio in
solution. PXRD patterns of the materials obtained on reaction of
Zn_MMO_P at molybdate concentrations of between 0.2 and
The crystalline phases identified in the materials obtained
following attempted reaction of precursor MMOs with ammo-
nium heptamolybdate or sodium or potassium molybdate solu-
tions are also summarised in Table 2. Co and Ni containing MMOs
showed little evidence of reactivity under any of the conditions
studied. PXRD patterns of the materials collected on reaction of
the MMOs remained unchanged from those of the precursor
MMOs. The use of lower calcination temperatures during MMO
preparation was not found to improve their reactivity. Zn and Cu
containing MMOs, in contrast, were reactive in all attempted
reactions. Fig. 3 shows the PXRD and FTIR analysis of the
precursor and product phases obtained for Zn. Corresponding
data for Cu may be found in the supplementary material. Reaction
of the MMOs in the presence of ammonium heptamolybdate
0
.025 M are shown in Fig. 4 (corresponding data for Cu may be
found in the supplementary material). The optimal concentration
for LTM formation was found to be 0.05 M for both Zn and Cu.
Assuming an LDH precursor composition of ideal formula
T
4
Al
2
(OH)12CO
3
2
ꢀ 2H O this would correspond to a T/Mo molar
ratio of 1. As the concentration of ammonium heptamolybdate
was increased a change in the phase selectivity of the reaction
was observed. New reflections appeared in the diffraction
patterns and those corresponding to the LTM phase decreased
in intensity. The sharp, low angle reflection at 2y=8.21 was
particularly noticeable. A search of the ICSD revealed that this
new phase bears a close resemblance (Fig. 4) to the family of
heteropolyoxomolybdates with the Anderson type structure
comprising the (H
cation and H O [33].
solutions yielded single phase
y
F LTM products. Changing the
source of molybdate to sodium or potassium molybdate led to a
change in the phase selectivity with reflections corresponding to a
reconstructed LDH phase appearing in the PXRD patterns. FTIR
analysis of the reconstructed LDH phases shows the reappearance
4
ꢁ
+
6
TMo
6
O
24
)
anion (where T=Zn or Cu), NH
4
2
ꢁ
1
2ꢁ
of a strong band at 1360 cm
indicating the presence of CO3
.
3.4. Reaction of MMO: influence of LDH precursor synthesis method
Based on the ideal compositions where a T/Mo ratio of 4 or 1.7
2
ꢁ
6ꢁ
would be expected for (MoO
4
)
or (Mo
7
O
24
)
containing LDHs
Fig. 5 compares SEM images taken of the LDH and MMO
precursors and the LTM products obtained during syntheses with
for a Zn/Al ratio of 2, respectively, the EDX analysis confirmed that

ARTICLE IN PRESS
2
02
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
Zn and Cu. Precursor LDHs prepared by coprecipitation exhibit the
features typically reported for this method; a wide particle size
distribution with aggregated particles which consist of small
significantly larger than those in the LTM prepared from
Zn_LDH_P.
(
typicallyo500 nm), plate-like primary crystallites [34].
Synthesis by the urea hydrolysis method yielded larger (average
particle size 2–3 m), plate-like crystallites, as expected [31]. No
3
.5. Reaction of MMO: influence of pH
m
The influence of pH on the reactivity of Zn–Al MMOs with
significant morphological changes were noted on thermal
decomposition of these LDHs to form their corresponding
MMOs. The corresponding LTMs formed on reaction of these
MMOs are found to have a similar average particle size to those of
their precursor LDHs although the LTM products appear
morphologically more well-defined. The LDH precursor
crystallite size clearly influences that of the LTM product, with
the LTM crystallites prepared from LDH precursor Zn_LDH_U
ammonium heptamolybdate and sodium molybdate solutions
0.05 M) is shown in Fig. 6. Addition of Zn_MMO_P to a solution of
ammonium heptamolybdate (Fig. 6(a)) resulted in an initial pH of
.4 and in the formation of a single phase LTM. Variation of the
initial pH between pH 5 and 10 had no significant influence on the
reactivity or selectivity of reaction. Reaction with sodium
molybdate solution (Fig. 6(b)) gave an initial pH of 9.0 and
(
5
F
y
resulted in the formation of
a reconstructed LDH phase.
Decreasing the pH to pH 5 led to a change in the phase
selectivity of the reaction. At this pH an LTM with the
structure was collected from solution.
F
x
3.6. LTMs: influence of preparative route
LTMs derived from MMO precursors, appear isostructural
to those prepared via solution precipitation. No significant
differences in the number, position or relative intensities of
reflections are observed between the PXRD patterns (Fig. 7)
or FTIR spectra of the products obtained. An SEM compari-
son (Fig. 8) reveals some differences in the microstructures
of the LTM products. Crystallites derived from MMO precursors
are larger and more uniform than those obtained by precipitation.
The particles sizes observed in the products derived from
MMO precursors also appear to be influenced by the particle
size of the LDH precursor, with larger average particle
sizes observed in Zn_LTM_MU products than in Zn_LTM_MP
products. An estimate of the crystallite sizes, s, of the LTM phases
obtained was calculated for comparison using a Scherrer based
analysis on the low angle (001) reflection. The results, which are
summarised in Table 3, support the observation of the formation
Fig. 4. PXRD analysis of the influence of molybdate concentration of the reaction
of Zn_MMO_P.
Fig. 5. SEM comparison of: LDH precursors (1) Zn_LDH_P, (2) Zn_LDH_U, (3) Cu_LDH_P; MMO intermediates (4) Zn_MMO_P, (5) Zn_MMO_U, (6) Cu_MMO_P; and LTM
products (7) Zn_LTM_MP, (8) Zn_LTM_MU and (9) Cu_LTM_MP.

ARTICLE IN PRESS
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
203
phases were found to have T/Al ratios between 1.6 and 1.9 which
is slightly lower than the stoichiometry of 2 present in solution
during LDH synthesis. As noted by Levin and co-workers for T=Zn
the T/Al ratio increased following reaction of the MMO with
ammonium heptamolybdate solution [23]. Noticeably the LTM
products, however, still contain significant amounts of Al.
Maximum loss of Al was observed on reaction of the LDH
precursor prepared by urea hydrolysis. The Al content of
Cu_LTM_MP obtained on reaction of Cu_MMO_P is comparable
to that of the MMO.
3.7. LTMs: characterisation of thermal properties
The TG profiles of Zn and Cu containing LTMs obtained by the
two different preparation routes and the PXRD patterns of the
residues remaining following their thermal analysis are shown in
Fig. 9. As Cu containing samples were found to be sensitive to
oxidation during thermal analysis (observed by a weight gain
upon heating in air, TG profiles shown in supplementary
2
material), the data presented were recorded under a N gas flow.
On heating the LTM, samples show a gradual weight loss at
temperatures below 200 1C. Above this temperature, all samples
undergo a sharp weight loss between 200 and 400 1C which can be
clearly identified in the corresponding DTG curves (see supple-
mentary material) and is thought to correspond to the simulta-
3 2
neous loss of NH and H O from the structure. A further weight
loss step, which is more noticeable for the Cu containing samples,
is observed above 700 1C. This loss is thought to be due to the
3
sublimation of MoO from the MMO phase which has been
previously observed during the thermal decomposition of other
Cu containing LTMs [22].
Analysis by FTIR and PXRD of the residues obtained following
thermal treatment combined with an in situ PXRD study provided
further insight into the nature of the LTM decomposition. Bands
þ
arising from the NH4 cations (e.g. the characteristic H–N–H
bending mode at 1410 cm ) present in the FTIR spectra of the as
Fig. 6. PXRD analysis of the influence of pH on the reaction of Zn_MMO_P with
ꢁ1
aqueous ammonium heptamolybdate (top) or sodium molybdate (bottom).
synthesised LTMs, remain visible in the residues obtained after
heating to 200 1C but are absent in the residues obtained
following heat treatment to 400 1C indicating loss of NH from
3
the structure. The in situ PXRD study confirmed the disappearance
of reflections corresponding to the LTM phase by 400 1C indicating
the simultaneous collapse of the LTM structure within this
temperature range (Fig. 10). The extrinsic reflections observed
in the two theta range of 40–451 are instrumental reflections
which arise as a result of diffraction from the high temperature
stage used.
Table 5 summarises the temperatures at which the principal
p
weight loss is observed, T , the total mass losses measured
following heat treatment to 400 1C, TML400 and the crystalline
phases identified following heat treatment to 800 1C. Ammonium
copper molybdate appears to be slightly more stable than
p
ammonium zinc molybdate (T =289 or 264 1C for T=Cu or Zn,
respectively). The TML400 values measured are in general slightly
higher than the expected 10.5% mass loss for an LTM of ideal
composition [12]. This is thought to be due to the presence of
physisorbed water in the as synthesised samples.
Although the form of the TG profiles is similar for LTM
products prepared by precipitation compared with those of LTMs
synthesised from MMO precursors, some differences may be
noted. As the temperature is increased, the LTM products derived
from MMOs show a larger initial weight loss (below 200 1C). This
weight loss is thought to be related to the loss of physisorbed
water from the sample and the differences observed might be a
consequence of the differences in sample morphologies between
the two samples (Fig. 8). Both products, Zn_LTM_MP and
Fig. 7. PXRD comparison of Zn_LTM_MP and Zn_LTM_P compared with the
simulated pattern of the
y
F structure.
of larger crystallites in LTM phases prepared via the MMO route
s=804 nm for Zn_LTM_MP compared with s=492 nm for
Zn_LTM_P).
Results of compositional analysis of the precursor and product
phases are reported in Table 4. The precursor LDH and MMO
(

ARTICLE IN PRESS
2
04
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
Fig. 8. SEM comparison of LTM products Zn_LTM_P (left) and Zn_LTM_MP (right).
Table 3
received continued interest. This property is exploited for the
decontamination of waste water streams [35,36], the preparation
of activated base catalysts [37,38], and for the synthesis of LDHs
containing various anionic species which may be difficult to
incorporate via other synthetic routes [39,40]. It is thought that,
depending on the extent of thermal decomposition and the
original LDH composition, the layer structure in the resulting LDH
phases remains largely unaltered from that of the precursor [26].
This permits manipulation of the interlayer species without major
variation of the stoichiometry or dispersion of the layer metal
cations.
Summary of the estimated crystallite sizes, s, of the LTM phases obtained,
calculated using a Scherrer based analysis for the (001) reflection with a shape
factor, k=0.9.
Sample
s (nm)
Zn_LTM_P
492
804
936
230
455
Zn_LTM_MP
Zn_LTM_MU
Cu_LTM_P
Cu_LTM_MP
By comparison exposure of MMOs to aqueous molybdate
containing solutions may result in several eventualities. It is
commonly associated with the preparation of (poly)oxomolyb-
date pillared LDH phases [41] in attempts to prepare novel
catalysts with improved properties [42–44]. Layered materials
have the potential to serve as a support for the catalytically active
species, the supported products acting as environmentally
friendly catalysts which can be conveniently recovered and
recycled [45] and may show increased catalytic activity and
selectivity compared with those of the unsupported catalysts [46].
LTMs have also been studied with a view to their use in similar
areas of catalysis as those of molybdate pillared LDHs e.g. for the
dehydrogenation of propane [47].
As the rehydration of MMOs in aqueous solution offers an
alternative, potentially greener, preparation route for the synth-
esis of both LTMs and molybdate containing LDHs it is important
to understand the distinctive phase selectivity which may result.
Fig. 11 summarises the possible outcomes both previously
reported and which have been observed during our
investigations. The route taken is dependent on several factors
including the MMO composition, molybdate source and reaction
conditions.
MMOs containing Ni or Co showed little evidence of reactivity
under the conditions studied. Ni and Co containing LDHs are
known to be difficult to reconstruct following thermal decom-
position. For Ni, reversibility is achievable if rehydration is
undertaken hydrothermally [48,49]. It was of interest to investi-
gate the influence of the presence of molybdate on the hydro-
thermal reaction of Ni containing MMOs. Fig. 12 shows the PXRD
patterns of the materials obtained following hydrothermal
reaction of Ni_MMO_P in the presence of ammonium, sodium
and potassium molybdate at 180 1C for 24 h. Changes in the
diffraction pattern were observed in all cases with broad, low
intensity reflections appearing in positions typical of those
expected for an LDH phase. No evidence of LTM formation was
observed under these conditions. Where Co is present
Table 4
Summary of compositional analyses of the LDH, MMO and LTM phases obtained
during synthesis by reaction of MMO precursor.
Sample
Wt% C
Wt% N
T:Mo
T:Al
Zn_LTM_P
Zn_LDH_P
0.0
1.9
0.7
0.0
2.2
0.1
0.1
0.1
2.5
1.3
0.1
4.9
0.0
0.0
2.3
0.0
0.0
2.4
2.6
0.0
0.0
2.6
0.9
–
–
1.7
1.8
2.5
1.6
1.9
4.0
–
Zn_MMO_P
Zn_LTM_MP
Zn_LDH_U
Zn_MMO_U
Zn_LTM_MU
Cu_LTM_P
Cu_LDH_P
Cu_MMO_P
Cu_LTM_MP
–
1.3
–
–
1.1
0.7
–
1.6
1.9
2.2
–
1.4
Cu_LTM_MP, also exhibit
measured values) than their counterparts prepared by
precipitation (Table 5).
The preparation method of the precursor is also found to
influence the stability of the resulting MMOs. MMOs derived from
T_LTM_MP precursors exhibit greater weight losses above 700 1C,
a higher stability (based on the
T
p
associated with increased sublimation of MoO
3
from their
structures, compared with those observed for MMOs derived
from T_LTM_P precursors. Analysis of the residues collected
following thermal treatment confirms differences in the nature of
the MMO derived from the precursor LTMs dependent on the
precursor preparation method (Table 5). Crystalline transition
4 3 2 9
metal oxides of TMoO and/or T Mo O structures were observed
to form following LTM decomposition depending on the precursor
LTM studied.
4
. Discussion
reconstruction may be possible if the formation of Co
3 4
O spinel
The ability of MMO phases, produced by the mild thermal
decomposition of LDHs, to reconstruct in aqueous solution has
is avoided during thermal decomposition by, for example,
calcining in a non-oxidising atmosphere [50]. With the desire to

ARTICLE IN PRESS
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
205
ꢁ
1
Fig. 9. Comparison of the TG profiles of (a) Zn_LTM_P, (b) Zn_LTM_MP, (c) Cu_LTM_P, (d) Cu_LTM_MP and (e) Zn_LTM_MU recorded under a N
The PXRD patterns of the residues remaining after their calcination to 800 1C are also shown.
2
gas flow of 30 ml min
.
solutions at room temperature and at 100 1C for 24 h. No
reaction was observed under any of the conditions studied.
Cu and Zn containing MMOs were reactive under all conditions
studied. The phase selectivity of the reaction was found to be
dependent on both the source and concentration of molybdate
species and on the reaction pH. Formation of LTMs possessing the
+
4
F
y
structure was achieved in the presence of NH
with a T/Mo
ratio in solution E1. This structure has only previously been
+
+
+
reported for A =NH
maximised in the
cation [19]. Formation of LTMs with the
4
. It is thought that A –layer interactions are
+
F
y
structure due to the symmetry of the NH
4
x
F structure in the
presence of sodium molybdate was only observed on reduction of
the reaction pH to 5.0. Above this pH LDH reconstruction occurred
preferentially. Analysis of the basal spacing and FTIR spectra of
2
ꢁ
the resulting LDH confirmed the incorporation of CO
3
and not
2
ꢁ
molybdate. The CO
3
anion is often incorporated into the LDH
structure if no precautionary measures are taken due to its high
abundance and strong interaction with the LDH interlayer [26].
Alternatively, the reaction time or molybdate concentration may
Fig. 10. In situ PXRD analysis of the thermal decomposition of Zn_LTM_MP.
not be sufficiently long or high for the incorporation of
ꢁ
(
polyoxo)molybdate species. Formation of terephthalate (TA )
Table 5
Summary of the total mass loss at 400 1C, TML400s, temperatures at which the
principal weight loss occurs during decomposition, T s, and the crystalline phases,
CPs, identified in resulting residues for the LTM products obtained.
intercalated LDHs by rehydration of MMOs, for example, is
ꢁ
2ꢁ
thought to occur in two stages; reconstruction of an OH =CO
containing LDH followed by anion exchange with TA . The extent
3
p
ꢁ
ꢁ
of anion exchange was dependent on the concentration of TA
Sample
T
p
(1C)
TML400 (%)
CPs
Mo
present [51]. Interestingly we have found that Anderson type
materials can also be produced by this Green Chemistry method
and note that they are also of catalytic interest.
Comparison of the LDH and LTM morphologies reveals
similarities, both in terms of particle size and morphology,
between the precursors and products of the MMO route. Levin
et al. commented on the close structural relationship between the
Zn_LTM_P
251
270
275
289
305
11.7
12.0
9.5
2
O
9
Zn
3
Zn_LTM_MP
Zn_LTM_MU
Cu_LTM_P
ZnMoO
ZnMoO
4
4
11.0
11.5
CuMoO
CuMoO
4
4
Cu_LTM_MP
3 2 9
and Cu Mo O
LDH and
y
F LTM structure [23]. This suggests that the mechanism
promote reactivity of Co_MMO_P precursors, Co containing LDHs
were calcined at 350 1C under a N environment for 1 h and the
resulting MMOs were reacted with molybdate containing
of LTM formation may be templated by the MMO particles in a
similar idea to the ‘memory effect’ for LDH reconstruction. The
mechanism of LDH reconstruction remains unclear. Initially it was
2

ARTICLE IN PRESS
2
06
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
Fig. 11. Summary of potential outcomes on reaction of MMO with aqueous molybdate containing solution.
depending on the conditions studied. This might explain such
conflicting arguments as that of ‘no morphological change during
reconstruction [56] with those who have shown growth of LDH
crystallites on the surface of initial hydrotalcite plates [53].
The concept of simultaneous dissolution and recrystallisation
with a templated crystal growth mechanism could explain not
only the morphological resemblance of the LTM products derived
from MMOs to their LDH precursors but also the differences in
observed phase selectivity of reaction. As the
x
F structure bears
less of a resemblance to that of the precursor LDH than the
F
y
structure, its formation would require more significant rearrange-
ment of the LDH structure. Lowering the reaction pH would
favour increased dissolution of the precursor MMO phase,
thus removing the presence of a templating phase and facilita-
ting structural rearrangement through dissolution and recrystal-
lisation.
Contrary to previous findings [23], compositional analyses of
the LTM products derived from MMOs indicate little change in the
quantity of Al present compared with that in the precursor LDHs.
Levin et al. related the reactivity of a MMO towards LTM
formation to the extent of Al dissolution during rehydration in
order to explain why MMOs possessing the rocksalt structure,
where the Al can remain in sites of octahedral coordination, are
generally thought to be less reactive [23]. Segregation of Al during
calcination and rehydration of LDHs have been reported by
several studies [57]. A mechanism solely involving dissolution
and recrystallisation, however, would not account for the
significant presence of Al which we observe. This Al is possibly
incorporated as an amorphous oxide phase or it could potentially
be present in the LTM structure either as a charge balancing
cation or occupying vacancies within the layer [16].
Fig. 12. Comparison of the PXRD patterns of the materials obtained following
exposure of Ni_MMO_P phases to aqueous solutions containing ammonium
heptamolybdate or sodium or potassium molybdate (Ni/MoE1).
The possibility of finding an Al on a lattice site has previously
been investigated by the attempted crystal structure refinement,
based on Rietveld analysis, of the ammonium zinc molybdate
structure. On refinement it was found that when a proportion of
Zn lattice sites were replaced with Al, the fractional occupancy
refined to zero indicating that this is not a likely possibility [23].
thought to occur topotactically [52] but recently convincing
evidence for the occurrence of dissolution and recrystallisation
during the rehydration process has been presented [53]. An in situ
PXRD study showed that reconstruction of an Mg–Al MMO
occurred rapidly with reflections corresponding to the LDH phase
appearing ‘almost instantaneously’ on exposure to aqueous
solution, even at room temperature [54]. In contrast, for the
reaction of MgO with an amorphous alumina phase a delay was
observed before the appearance of LDH reflections which was
thought to be the consequence of an additional diffusion/mixing
step required for LDH formation in that case [55]. It is possible
that the two processes occur simultaneously to varying extents
2
7
We applied
Al solid state NMR (not shown) to study the
3
+
conversion of the Al
between octahedral and tetrahedral
environments during calcination of the LDH precursors and on
reaction of the resulting MMOs with ammonium heptamolybdate.
On calcination we observed the expected conversion from
chemical shifts of octahedral to tetrahedral coordination, corre-
sponding to previously reported observations. Following reaction
of the MMO with ammonium heptamolybdate a change in the
NMR spectra is also observed, in this case with conversion
from tetrahedral to octahedral coordination. Peaks due to
3
+
both tetrahedral and octahedral Al
remain in the LTM

ARTICLE IN PRESS
S. Mitchell et al. / Journal of Solid State Chemistry 183 (2010) 198–207
207
phases indicating that the Al is in more than one coordination
environment.
[11] R.Z. Ma, Z.P. Liu, K. Takada, N. Iyi, Y. Bando, T. Sasaki, J. Am. Chem. Soc. 129
2007) 5257–5263.
12] L.A. Palacio, A. Echavarria, D.A. Hoyos, C. Saldarriaga, Solid State Sci. 7 (2005)
043–1048.
[13] A. Clearfield, A. Moini, P.R. Rudolf, Inorg. Chem. 24 (1985) 4606–4609.
(
[
We also studied the homogeneity of the LTM compositions by
EDX analysis of a range of samples. This showed that there is little
sample heterogeneity (areas with distinct chemical compositions)
within the resolution limits of the electron microscope. This
conclusion was supported by observation of the back scattered
electron images which showed little phase contrast between
particles. From these results we can conclude that if phase
separation occurs during reaction of the MMO, for example if the
Al is present as an amorphous alumina component, this must
occur at a nanoscopic level.
1
[
[
14] M. Jang, T.J.R. Weakley, K.M. Doxsee, Chem. Mater. 13 (2001) 519–525.
15] D.Q. Chu, C.L. Pan, L.M. Duan, J.Q. Xu, L.M. Wang, J.H. Yu, T.G. Wang, Inorg.
Chem. Commun. 5 (2002) 989–992.
[16] C.D. Wu, C.Z. Lu, X. Lin, S.F. Lu, H.H. Zhuang, J.S. Huang, J. Alloys Compd. 368
(2004) 342–348.
[
[
17] H. Pezerat, I. Mantin, S. Kovacevi, Compt. Rend. 262 (1966) 95–98.
18] L.A. Palacio, A. Echavarria, C. Saldarriaga, J. Inorg. Mater.
3
(2001)
367–371.
[19] D. Levin, S.L. Soled, J.Y. Ying, Inorg. Chem. 35 (1996) 4191–4197.
[
[
[
[
20] S. Soled, D. Levin, S. Miseo, J. Ying, Stud. Surf. Sci. Catal. 118 (1998)
59–367.
21] A. Moini, P.R. Rudolf, A. Clearfield, J.D. Jorgensen, Acta Crystallogr. C 42 (1986)
1667–1669.
3
Differences in the thermal properties of LTM products might
be expected due to the presence of Al in those prepared from
MMOs. Cu containing LTMs are found to have a slightly improved
thermal stability compared to Zn containing TMMs and this may
22] L.A. Palacio, A. Echavarria, C. Saldarriaga, Stud. Surf. Sci. Catal. 135 (2001)
2888.
23] D. Levin, S.L. Soled, J.Y. Ying, Chem. Mater. 8 (1996) 836–843.
indicate a greater stability of the
F
y
structure with Cu.
[24] A. Clearfield, M.J. Simms, R. Gopal, Inorg. Chem. 2 (1976) 335–338.
[
[
25] H. Pezerat, Compt. Rend. 261 (1965) 5490.
26] J. He, M. Wei, B. Li, Y. Kang, D.G. Evans, X. Duan, in: X. Duan, D.G. Evans (Eds.),
Layered Double Hydroxides (Structure and Bonding), Springer, Berlin, 2006,
pp. 89–119.
5
. Conclusions
Several possible reaction pathways exist for the reaction of
[27] A. de Roy, C. Forano, J.P. Besse, in: V. Rives (Ed.), Layered Double
Hydroxides: Present and Future, Nova Science Publishers Inc., New York,
2001, pp. 1–40.
MMOs derived from LDH precursors in aqueous molybdate
containing solutions. MMOs of different compositions differ in
their reactivity. For a given MMO the phase selectivity of the
reaction is dependent on factors such as the nature of the
[
[
28] M.T. Pope, Heteropoly and Isopoly Oxometalates, Springer, New York, 1983.
29] T. Hibino, A. Tsunashima, Chem. Mater. 9 (1997) 2082–2089.
[30] M. Turco, G. Bagnasco, U. Costantino, F. Marmottini, T. Montaneri, G. Ramis,
G. Busca, J. Catal. 228 (2004) 43–55.
[31] U. Costantino, F. Marmottini, M. Nocchetti, R. Vivani, Eur. J. Inorg. Chem. 10
+
countercation A , present in solution, the molybdate concentra-
(1998) 1439–1446.
tion and the reaction pH. Although LTMs prepared from MMOs
appear isostructural to those obtained by precipitation synthesis
they are found to have different compositions, morphologies and
thermal properties. The preparation route of LDH precursors was
also shown to influence both the crystallite size and thermal
properties of the LTM products obtained.
[32] F. Kooli, C. Depege, A. Ennaqadi, A. de Roy, J.P. Besse, Clays Clay Miner. 45
(1997) 92–98.
[33] C.C. Allen, R.C. Burns, G.A. Lawrance, P. Turner, T.W. Hambley, Acta
Crystallogr. C 52 (1997) 7–9.
[34] Z.P. Xu, G. Stevenson, C.Q. Lu, G.Q. Lu, J. Phys. Chem.
16923–16929.
B 110 (2006)
[
[
35] L.A. Lv, Y. Wang, M. Wei, H.J. Cheng, J. Hazard. Mater. 152 (2008) 1130–1137.
36] M. Bouraada, M. Lafjah, M.S. Ouali, L.C. de Menorval, J. Hazard. Mater. 153
(2008) 911–918.
[
[
[
37] S. Abello, F. Medina, D. Tichit, J. Perez-Ramirez, J.C. Groen, J.E. Sueiras,
P. Salagre, Y. Cesteros, Chem.-A Eur. J. 11 (2005) 728–739.
38] X.D. Lei, F.Z. Zhang, L. Yang, X.X. Guo, Y.Y. Tian, S.S. Fu, F. Li, D.G. Evans,
X. Duan, A.I.Ch.E. J. 53 (2007) 932–940.
Acknowledgments
The authors would like to acknowledge EPSRC and CSIC for
funding. Morgane Presle and Aur e´ lie Guenet, visiting students to
the group, are thanked for their preliminary work in this area.
39] P. Kovar, K. Melanova, V. Zima, L. Benes, P. Capkova, J. Colloid Interface Sci.
3
19 (2008) 19–24.
[40] C.A.S. Barbosa, A.M.D.C. Ferreira, V.R.L. Constantino, Eur. J. Inorg. Chem.
2005) 1577–1584.
(
[
[
[
[
41] K. Chibwe, W. Jones, Chem. Mater. 1 (1989) 489–490.
42] A.L. Maciuca, E. Dumitriu, F. Fajula, V. Hulea, Appl. Catal. A 338 (2008) 1–8.
43] D. Carriazo, C. Martin, V. Rives, Catal. Today 162 (2007) 153–161.
44] P. Atkins, C.L. Depege, C. Forano, M. Rene, US Patent 6124506, 2000.
Appendix A. Supplementary material
Supplementary data associated with this article can be found
in the online version at doi:10.1016/j.jssc.2009.10.011.
[45] B.M. Choudary, S. Madhi, N.S. Chowdari, M.L. Kantam, B. Sreedhar, J. Am.
Chem. Soc. 124 (2002) 14127–14136.
[
[
46] E. Gardner, T.J. Pinnavaia, Appl. Catal. A 167 (1998) 65–74.
47] D. Levin, J.Y. Ying, Stud. Surf. Sci. Catal. 110 (1997) 367–373.
References
[48] F. Kooli, V. Rives, M.A. Ulibarri, Inorg. Chem. 34 (1995) 5114–5121.
[
[
[
[
49] P. Benito, I. Guinea, F.M. Labajos, V. Rives, J. Solid State Chem. 181 (2008)
87–996.
50] J. Perez-Ramirez, G. Mul, F. Kapteijn, J.A. Moulijn, Mater. Res. Bull. 36 (2001)
767–1775.
51] E.L. Crepaldi, J. Tronto, L.P. Cardoso, J.B. Valim, Colloids Surf. A 211 (2002)
03–114.
52] T. Sato, H. Fujita, T. Endo, M. Shimada, A. Tsunashima, React. Solids 5 (1988)
19–228.
9
[
[
[
1] M.M. Barsan, C.F. Thyrian, Catal. Today 81 (2003) 159–170.
2] M. Honda, K. Tanaka, I. Watambe, DE Patent 2112938 19720127, 1972.
3] M. Breysse, C. Geentet, P. Afanasiev, J. Blanchard, M. Vrinat, Catal. Today 130
1
(
2008) 3–13.
4] L. Barrio, J.M. Campos-Martin, M.P. de Frutos, J.L.G. Fierro, Ind. Eng. Chem. Res.
7 (2008) 8016–8024.
1
[
4
2
[
[
[
5] J.M. Mitchell, N.S. Finney, J. Am. Chem. Soc. 123 (2001) 862–869.
6] J.S. Chung, R. Miranda, C.O. Bennett, J. Catal. 114 (1988) 398–410.
7] I.D. Ospina, L.A. Palacio, A. Echavarria, A. Simon, C. Saldarriaga, Micropor.
Mesopor. Mater. 47 (2001) 303–309.
[
[
[
53] T.S. Stanimirova, G. Kirov, J. Mater. Sci. Lett. 20 (2001) 453–455.
54] F. Millange, R.I. Walton, D. O’Hare, J. Mater. Chem. 10 (2000) 1713.
55] S. Mitchell, T. Biswick, W. Jones, G. Williams, D. O’Hare, Green Chem. 9 (2007)
3
73–378.
56] F. Delorme, A. Seron, M. Bizi, V. Jean-Prost, D. Martineau, J. Mater. Sci. 41
2006) 4876–4882.
57] T. Hibino, A. Tsunashima, Chem. Mater. 10 (1998) 4055–4061.
[
[
8] D.Q. Chu, C.L. Pan, L.M. Duan, J.Q. Xu, L.M. Wang, J.H. Yu, T.G. Wang, Inorg.
Chem. Commun. 5 (2002) 989–992.
9] S. Lei, K. Tang, Q. Liu, Z. Fang, Q. Yang, H. Zheng, J. Mater. Sci. 41 (2006)
[
[
(
4
737–4743.
[
10] J.S. Xu, D.F. Xue, Y.C. Zhu, J. Phys. Chem. B 110 (2006) 17400–17405.